Lightweight lexical augmentation for robust transformer-based student feedback classification

Key Points

• Improved classification accuracy demonstrates the effectiveness of lexical augmentation in transformers.
• An increase of 15% in classification accuracy highlights enhanced robustness in feedback processing.
• Observational analysis using transformer models evaluated enhancement in student feedback classification.
• The findings may enable better understanding of student insights, but further testing in diverse settings is needed.